'AK Mods' SC Class A 15W Amplifier

[Spirit]

Jan 28, 2012 6:48:24 GMT @sandyk said:

Will's Class A PCBs incorporate the offset corrector as in the HA/preamp.
Alex

Shows how much attention I pay
I think I have some JLHs floating around from Greg's run.
Though I couldn't see them last time I looked.
Can't remember if Jeff ended up dropping them off...

Cheers,
Phil

[jeffc]

I think I have some JLHs floating around from Greg's run.
Though I couldn't see them last time I looked.
Can't remember if Jeff ended up dropping them off...

Hi Phil,

All of the ones sent by Alex I stole and have used except 1 also already stuffed. You can tell who your friends are . Should have got more of Will's single PCBs in the last GB with the PK DAC. I got 6 but have stuffed 4 already for the 15W Class A and sent 1 to Alex for the same for close positioning next to the amp PCBs, so no go there either.

Maybe if enough interest, Will can organize another JLH PCB lot along with the amp and PSU PCBs.

Hi Shaun, nice lookin' amp and PSU boxes on display there. And terrific to see how you've approached it, as there is lots to consider, and the pics offer plenty of tips, Well Done.

cheers.. jeffc

[Deleted]

You can tell who you're friends are

Phil
If you are unable to obtain the single JLHs for close positioning with the main PCBs, let me know and I will pop a couple of my remaining G.E. designed duals in the mail for you.You can always use a similar layout to Shaun.
Regards
Alex
P.S.
I have reserved a couple of coil bobbins for both you and Jeff.

[Spirit]

Jeff -
At least they're being put to use, I guess
I do vaguely recall saying I was unlikely to use them and you might as well hang on to them

Alex, do you mean for the zobel network?
I've still got the bobbins from the SCHA left over, so can use those presumably?
I'll wait and see if there's enough demand for JLHs in the GB, otherwise a couple of duals would be great, thanks

Cheers,
Phil

[Deleted]

I've still got the bobbins from the SCHA left over, so can use those presumably?

Hi Phil
The bobbins from the SC HA even have the correct number of turns,(4.7uH) but if your speakers dip below 8 ohms I would recommend using .8mm enamelled wire instead.
Alex

[Deleted]

Updated Schematic.

Q108 has a 120 ohm emitter resistor.
Q103 has a 200 ohm emitter resistor.
Q101 and Q102 have 200 ohm emitter resistors.
The 1N5819 has been replaced by a SF12 diode.
The voltage readings are only a guide, as they were taken with the cover off,and the balance adjustment was not correct due to this.
Alex

P.S.
Will's new PCB may have different component numbering .
Also, in cold climates, if front end balance can't be readily achieved ,we may need to revert back to using the 1N5819 Schottky Diode due to it's lower forward voltage, and put up with touchier set up of the front end balance.
If the difference obtained is just short of achieving complete balance, then replacing Q108's 120 ohm emitter resistor with a 130 ohm 1% MF resistor MAY get you there without reverting back to the 1N5819.

Attachments:

[jeffc]

Phil
If you are unable to obtain the single JLHs for close positioning with the main PCBs, let me know and I will pop a couple of my remaining G.E. designed duals in the mail for you.You can always use a similar layout to Shaun.
Regards
Alex
P.S.
I have reserved a couple of coil bobbins for both you and Jeff.

cheers Alex, you're a top bloke

And guys, if any of you are struggling to locate heat sinks for the BC337/BC327 driver transistors, you can find some here.
www.ebay.com.au/itm/TO-18-Heatsink-Small-Ideal-QRP-Transmitters-5-PK-Good-Price-Hard-Find-/290659529333?pt=LH_DefaultDomain_0&hash=item43aca87275
I got 4 sets of 5 as that's all was advertised as remaining but the stash is obviously larger as more are now being advertised. Alex you have a couple of sets on the way and Phil, I have set of 4 for you too to help make up for those stolen JLH PCBs.

cheers.. jeffc

[Deleted]

And guys, if any of you are struggling to locate heat sinks for the BC337/BC327 driver transistors, you can find some here.http://www.ebay.com.au/itm/TO-18-Heatsin....=item43aca87275

As the driver transistors don't get extremely hot, you could most likely even get away with clipping on 2 x M205 PCB mounting fuse clips on each device if you get stuck.. Just open them up slightly.

[Deleted]

Possible Voltage Regulator failure in Offset Corrector
There have now been a couple of reports of voltage regulator failure , both in the preamp and 15W Class A.
There isn't anything in the Offset Corrector circuit (attached) that can cause this.It may be due to particular brands of 78Lxx devices,
so if room permits and the PCB holes are large enough diameter, it may be better to use the 7815 and 7915.It is also advisable to fit 8.2 ohm PW5 resistors at the output until the amplifier has been verified to be working correctly. The Loudspeaker protector should be used at all times to prevent accidental damage, as DC coupled amplifiers are unforgiving of faults.
I found the attached explanation in an article about a general purpose low voltage PSU in Silicon Chip magazine for March 2011, where they fitted reverse connected diodes across the outputs of their voltage regulators.The author was N.Vinen who designed the S.C. DAC.
Alex

The diodes protect the circuit in case of asymmetric loads.
Such loads can pull the positive rail negative, or the negative rail positive during switch-off or over current conditions
D5 and D6 clip these transient voltages and prevent damage to the regulators and filter capacitors under such conditions .
D5 and D6 overcome the boot-strapping problems that can occur with certain brands of regulators (mainly 78Lxx types)

Attachments:

[jeffc]

There have now been a couple of reports of voltage regulator failure , both in the preamp and 15W Class A.

Alex, hope you included me on that list as I've killed some and nice to know the likely reason why.

cheers.. jeffc

[Deleted]

Hi Jeff
I did remember that you had the same problem initially.
When using the TL071 the maximum current drawn from the regulators should be no more around 3mA, so it's not as if the 100mA rated Voltage regulators are working very hard.
Frans is not in agreement with the Silicon Chip explanation, believing that the damage would be more is more likely to occur at shut down.
Frans would use the 7815 and 7915 instead of the 78L15 and 79L15
and not worry about the protective diodes in this case.
That would be my preferred option too, PCB permitting.
Alex

[Deleted]

Hi Alex,
with regard to input and output voltages through my LT Reg/JLH PSU combo board, I'm not seeing much of a drop in voltage if any after the JLH section.
Output point 1..... LT/reg positive section out voltage 20.2
Output point 2.....JLH positive section out voltage hovers around 20.2/20.1

Negative section drops .1 of a volt steady
Should the JLH out voltage drop the full .2v to 20 volts?
I've not run under load yet as heat sinking arrangements are not yet in place.
Regards Alan

[Deleted]

Hi Alan
Is this with the CL section in circuit, and a 1 ohm emitter resistor for the 2SA1930 in the +VE rail, and 2SC5171 in the -VE rail Current Limiters? If the CL section is bypassed then you will see no difference in voltage.
Try loading the outputs with a PW5 resistor from 100 ohms to 470 ohms that you may have on hand, and see what voltage drops you get then.

Regards
Alex

[Deleted]

Frans is not in agreement with the Silicon Chip explanation, believing that the damage would be more is more likely to occur at shut down.Frans would use the 7815 and 7915 instead of the 78L15 and 79L15and not worry about the protective diodes in this case.

Now seeing the schematic and noticing it has 100uF caps on the output side AND the current consumption of the opamp is low AND taking in consideration the amp is class-A and thus and switch off probably may drop below the voltage on the output of the regs that in that case can be higher MAY result in damage to the regs. In this case 100uF on the output and a rapid fall of the input voltage reverse diodes are recommended.
As Alex mentioned the damage would occur on switch off and will be noticed on start-up the next time IF this is the case as only in these circumstances the latch-up can occur.

recommendations are lower the output caps in value to anything between 1uf to max 10uF and or fit the diodes.
My experience with the TO-92 L type regs is they oscillate easily when not decoupled properly with 100nF ceramics as close as possible to the output and ground pin and at least 330nF ceramic as close as possible to the input pin and ground.

It may also simply be a dodgy part or perhaps they have been soldered too long and have had some latent damage in case it breaks down during operation.

Personally because of disappointing results I prefer normal or M type (TO-220) type regs over the L types.

[Deleted]

Feb 1, 2012 22:44:43 GMT @solderdude said:

Frans is not in agreement with the Silicon Chip explanation, believing that the damage would be more is more likely to occur at shut down.Frans would use the 7815 and 7915 instead of the 78L15 and 79L15and not worry about the protective diodes in this case.

Now seeing the schematic and noticing it has 100uF caps on the output side AND the current consumption of the opamp is low AND taking in consideration the amp is class-A and thus and switch off probably may drop below the voltage on the output of the regs that in that case can be higher MAY result in damage to the regs. In this case 100uF on the output and a rapid fall of the input voltage reverse diodes are recommended.
As Alex mentioned the damage would occur on switch off and will be noticed on start-up the next time IF this is the case as only in these circumstances the latch-up can occur.

recommendations are lower the output caps in value to anything between 1uf to max 10uF and or fit the diodes.
My experience with the TO-92 L type regs is they oscillate easily when not decoupled properly with 100nF ceramics as close as possible to the output and ground pin and at least 330nF ceramic as close as possible to the input pin and ground.

It may also simply be a dodgy part or perhaps they have been soldered too long and have had some latent damage in case it breaks down during operation.

Personally because of disappointing results I prefer normal or M type (TO-220) type regs over the L types.

Hi Guys

thanks for the info and yes the popping 78L reg was me

BUT

i think i have managed to solve the problem,

remember how confident i was that the protection diodes on the Positive regs where not needed . well the switch off managed to kill one decoupling cap on the amp board and put a nice big bulge in the 100uf cap on the other reg.

so the data sheet says protection diodes on the positive reg are NOT needed BS. well they are if you value your caps ;D ;D

i have protection diodes on my negative regs (data sheet says they are needed) with no problems.

just in case anyone else may be stupid enough to believe everything they read. (i was).

so i think it may be more of a stupidity issue than the fault of the 78L15 regs

so i just have the high DC offset to deal with which seems to be identical on both boards.

all good clean fun

take care

[Deleted]

Hi Alex,
thanks for the reply,

Is this with the CL section in circuit, and a 1 ohm emitter resistor for the 2SA1930 in the +VE rail, and 2SC5171 in the -VE rail Current Limiters?

Yes CL as above.
I'd built up the two PSU boards with LT1085 and LT1033 Regs and positive/negative JLH sections combined on each..........quite neat at 105mm X 135mm,.....but only if the bloody thing works
had a few senior moments with the voltages until I realised that the DMM was playing up and dumped it for another.
However by that point I'd scrutinised schematic....layout and board..........every which way but loose............and found I'd the BC639s reversed in the positive sections.
Replaced the backward buggers....fired up and had joy..........
So I'll load the outputs and see what transpires.
Thanks Alex,

Chris,
I replaced the BC639s in the four supply modules running Will's PK and to my great pleasure think the SQ is mucho better
What do I mean.......think?.........of course it is those last three files ;)really pull out all the stops.
Regards,
Alan

[Deleted]

Hi Alan
Don't forget that the PSU with the CL in circuit can't supply the output stage's 1A requirements.It's only suitable for the front end of the amplifier.
Regards
Alex

[Deleted]

Feb 2, 2012 0:22:18 GMT drymdrum said:

Hi Alex,
thanks for the reply,

Is this with the CL section in circuit, and a 1 ohm emitter resistor for the 2SA1930 in the +VE rail, and 2SC5171 in the -VE rail Current Limiters?

Yes CL as above.
I'd built up the two PSU boards with LT1085 and LT1033 Regs and positive/negative JLH sections combined on each..........quite neat at 105mm X 135mm,.....but only if the bloody thing works
had a few senior moments with the voltages until I realised that the DMM was playing up and dumped it for another.
However by that point I'd scrutinised schematic....layout and board..........every which way but loose............and found I'd the BC639s reversed in the positive sections.
Replaced the backward buggers....fired up and had joy..........
So I'll load the outputs and see what transpires.
Thanks Alex,

Chris,
I replaced the BC639s in the four supply modules running Will's PK and to my great pleasure think the SQ is mucho better
What do I mean.......think?.........of course it is those last three files ;)really pull out all the stops.
Regards,
Alan

Hi Alan

well done

it's onwards and errr onwards

take care

[Deleted]

HI Alex,

Don't forget that the PSU with the CL in circuit can't supply the output stage's 1A requirements.It's only suitable for the front end of the amplifier.

Yes I've noted that Alex and have two output points for both positive and negative supply on each board. Point 1 @20.2v before the CL/JLH and Point 2 after the JLH @20v.
Regards,
Alan

[Deleted]

DRYMDRUM's PSU for The AK Mods SC 15 Watt Amp
What I t Looks Like

[jeffc]

Hi Alan,

Those PSU PCBs are a "work of art" and terrific idea using old computer SMPS cases for transformer shielding . I've got heaps of these lying about . Well done!!

cheers.. jeffc

[Deleted]

Hi Alan
Very innovative, but remember that to get full benefit from the JLHs , they should have fairly short leads to the Class A PCBs.
Kind Regards
Alex

[Deleted]

Hi Alex,
as the PSU boards will be in the main amp case the JLH supply will be right up the Khyber Pass............if you will pardon the expression.
I just hope that innovative gets fair lady.
Spent a few hours on my main amp board layout tonight...........very encouraged with progress looks like I'll be cutting copper soon.
However not this weekend as we are off tomorrow.........shhhh...it today for a curling weekend at Kinross I'll probably be knackered on Monday and take a week to recover.
Regards,
Alan

[Deleted]

Feb 2, 2012 14:15:44 GMT drymdrum said:

DRYMDRUM's PSU for The AK Mods SC 15 Watt Amp
What I t Looks Like

Hi Alan

grrrrreat work on those boards they look lovely.

take care Hifi Superman

[Deleted]

Chris,
I replaced the BC639s in the four supply modules running Will's PK and to my great pleasure think the SQ is mucho better

Alan,

Confirmed here too

Your amp build is looking very impressive as we have come to expect